The P histo-blood group system is the last of the known carbohydrate defined blood group systems for which the molecular genetic basis has not yet been clarified. The P blood group system involves two major blood group pheno-types, P1+ and P1− with approximate frequencies of 80 and 20%, respectively (Landsteiner and Levine, 1927; Daniels et al., 1999). P1− individuals normally express the P antigen (P1− is designated P2 when P antigen expression is demonstrated), but the rare Pk phenotype lacks the P antigen, while the rare p phenotype lack both P and Pk antigens (for reviews see (Watkins, 1980; Marcus, 1989; Marcus and Kundu, 1980; Issitt and Anstee, 1998; Bailly and Bouhors, 1995)). The P1+ phenotype is defined by expression of the neolacto-series glycosphingolipid P1 (for structures see Table I) (Naiki et al., 1975).
TABLE IStructures of glycosphingolipids referred to inthis studyaP blood groupStructureantigenCDH, LacCerGalβ1-4Glcβ1-1CerpCTH, Gb3Galα1-4Galβ1-4Glcβ1-1CerPkGlobosideGalNAcβ1-3Galα1-4Galβ1-4Glcβ1-1CerPSialyl-Gal-GlobosideNeuAcα2-3Galβ1-3GalNAcβ1-3Galα1-4Galβ1-4Glcβ1-1CerLKEParagloboside, PGGalβ1-4GlcNAcβ1-3Galβ1-4Glcβ1-1CerP1Galα1-4Galβ1-4GlcNAcβ1-3Galβ1-4Glcβ1-1CerP1aKey: CDH, ceramide dihexoside (lactosylceramide, LacCer); CTH, ceramide trihexoside (Gb3, globotriaosylceramide); globoside, Gb4 (globotetraosylceramide); Cer, ceramide; Gal, D-galactose; Glc, D-glucose; GalNAc, N-acetyl-D-galactosamine; GlcNAc, N-acetyl-D-glucosamine; NeuAc, N-acetylneuraminic acid.
In contrast, the P, Pk, and p antigens constitute intermediate steps in biosynthesis of globo-series glycolipids and give rise to P1k, P2k, and p phenotypes (Naiki and Marcus, 1974). While the rare Pk phenotype show the same frequency of P1 antigen expression as individuals expressing the P antigen, the p phenotype is always associated with lack of P1 antigen expression. Extensive studies of the chemistry, biosynthesis, and genetics of the P blood group system identified the antigens as being exclusively found on glycolipids, with the blood group specificity being synthesized by at least two distinct glycosyltransferase activities; UDP-galactose: β-D-galactosyl-β1-R 4-α-D-galactosyltransferase (α4Gal-T) activity(ies) for Pk and P1 syntheses and UDP-GalNAc: Gb3 3-β-N-acetylgalactosaminyltransferase activity (EC 2.4.1.79) for P synthesis [for reviews see (Issitt and Anstee, 1998; Bailly and Bouhors, 1995)]. At least two independent gene loci, P and P1Pk, are involved in defining these antigens. The P blood group associated LKE antigen shown to be the extended sialylated Gal-globoside structure (Tippett et al., 1986), may involve polymorphism in an α2,3sialyltransferase activity.
A longstanding controversy has been whether a single or two independent α1,4galactosyltransferases catalyze the synthesis of the P1 neolacto-series glycolipid antigen and the Pk globo-series structure (Watkins, 1980; Marcus, 1989; Marcus and Kundu, 1980; Issitt and Anstee, 1998; Bailly and Bouhors, 1995). Several hypotheses have been proposed, including: i) a model with two distinct functional genes being allelic or non-allelic, where the P1 gene encodes a broadly active α4Gal-T, the Pk gene encodes a restricted α4Gal-T, and a null allele encodes a non-functional protein; ii) a model with two distinct non-allelic genes, where P1 encodes an α4Gal-T that can only synthesize P1 structures and the Pk encodes an α4Gal-T that only synthesize the Pk structure; and iii) a model where one gene locus encodes an α4Gal-T that is modulated by an independent polymorphic gene product to synthesize both P1 and Pk structures. Bailly et al. (Bailly et al., 1992) reported that kidney microsomal α4Gal-T activity from P1 individuals does not compete for the two substrates used by P1 and Pk α4Gal-T activities, and no accumulative effect in P1 synthase activity was observed when mixing microsomal fractions from individuals of P1 and pk groups. Based on this Bailly and colleagues suggested the existence of two distinct genes, coding for one P1 α4Gal-T with exclusive activity for neolacto-series substrates and one Pk α4Gal-T with exclusive activity for the globo-series substrate. Since p individuals lack the P1 antigen this model inferred that two independent genetic events inactivating both genes was responsible for the p phenotype.
Several approaches to gain insight into the P blood group α4Gal-T gene(s) have been attempted. Purification of the mammalian enzymes has not been successful, but identification and cloning of a bacterial α4Gal-T involved in lipopolysaccharide biosynthesis (Gotschlich, 1994; Wakarchuk et al., 1998) potentially provided a strategy to clone the mammalian genes using sequence similarity. Previously, a bacterial α3fucosyltransferase was identified in helicobactor pylori using a short sequence motif conserved among mammalian α3fucosyltransferases (Martin et al., 1997). BLAST analysis of gene databases with the coding region of the α4Gal-T gene from Neisseria Meningococcae resulted in identification of two human genes encoding putative type II transmembrane proteins with low sequence similarity to the bacterial gene1. The genes have open reading frames encoding 349 (EST cluster Hs.251809) and 371 (EST cluster Hs.82837) amino acid residues, and are located at 8q24 and 3p21.1, respectively. Previously, we established Epstein-Barr virus transformed B cells from two p individuals (Wiels et al., 1996). Only the gene at 3p21.1 was found to be expressed in the EBV-transformed p cells, as well as in Ramos cells known to have high Pk α4Gal-T activity. Sequencing of the coding region of the gene showed no mutations in p cells. Finally, expression of full coding or truncated, secreted constructs of either gene in insect cells failed to demonstrate glycosyltransferase activity with a large panel of substrates, including lactosylceramide, for Pk α4Gal-T activity. 1 R. Steffensen, J. Wiels, E. P. Bennett, and H. Clausen, unpublished observation.
Access to the Pk α4Gal-transferase gene would allow production of efficient enzymes for use in galactosylation of glycosphingolipids, oligosaccharides, and glycoproteins. Such enzymes could be used, for example, in pharmaceutical or other commercial applications that require enzymatic galactosylation of these or other substrates in order to produce appropriately glycosylated glycoconjugates having particular enzymatic, immunogenic, or other biological and/or physical properties. The P blood group system is implicated in important biological phenomena. Blood group p individuals have strong anti-P1PPk IgG antibodies and these are implicated in high incidence of spontaneous abortions (Yoshida et al., 1994). The globoseries glycolipid antigens constitute major receptors for microbial pathogens with the Galα1-4Gal linkage being an essential part of the receptor site (for a review see (Karlsson, 1998)). The Pk glycolipid is the CD77 antigen, a B cell differentiation antigen, which is able to transduce a signal leading to apoptosis of the cells (Mangeney et al., 1993). Furthermore, the association of this glycolipid with the type I interferon receptor or with the HIV-1 co-receptor, CXCR4, seems to be crucial for the functions of these receptors (Taga et al., 1997; Puri et al., 1999). Cloning of the Pk synthase is an important step toward understanding the biological roles of the globo-series class of glycolipids, and a first step in elucidating the molecular genetics of the P blood group system. Availability of the Pk synthase gene is important for elucidating the many biological roles of the globo-series class of glycolipids, and may offer new avenues for diagnostic and therapeutic measures. Consequently, there exists a need in the art for UDP-galactose: β-D-galactose-R 4-α-D-galactosyltransferase and the primary structure of the gene encoding this enzyme. The present invention meets this need, and further presents other related advantages, as described in detail below.